TLDR
Root cause analysis is best understood through real scenarios. This article walks through five manufacturing case studies: a dimensional nonconformance, a repeat customer complaint, a machine breakdown, an incoming material failure, and an audit finding. Each example shows the investigation process from problem identification through corrective action.
These case studies demonstrate how structured investigation methods work in practice and illustrate the difference between surface-level fixes and systemic corrective actions.
Learning from Real Investigations
Root cause analysis techniques are straightforward to describe in the abstract. The challenge is applying them to messy, real-world problems where data is incomplete, timelines are compressed, and multiple factors are in play. The following five case studies are drawn from common manufacturing scenarios. Each one follows the investigation from initial problem through root cause identification and corrective action.
Case Study 1: Dimensional Nonconformance on a Machined Component
The Problem
A Tier 2 automotive supplier receives a customer rejection for a machined aluminum housing. The bore diameter on 340 parts measures 25.12 to 25.18 mm against a specification of 25.00 +/- 0.05 mm. All rejected parts were produced during second shift on a single CNC machining center over a two-day period.
Containment
The quality team quarantines all second-shift production from the affected machine for the two-day window. In-transit inventory is placed on hold at the customer’s receiving dock. An expedited 100% sort recovers 340 nonconforming parts and confirms first-shift production during the same period is within specification.
Investigation
The team uses a 5 Whys analysis.
Why 1: Why is the bore oversized? The cutting tool is removing less material than programmed, leaving the bore larger than nominal. (Note: for a bore, removing less material leaves the diameter smaller. The team verifies the failure mode. In this case, the tool was worn and deflecting, creating an oversized bore on one side.)
Why 2: Why was the tool worn beyond its effective life? The tool had been in use for 1,800 parts against a recommended change interval of 1,200 parts.
Why 3: Why was the tool not changed at 1,200 parts? The tool change counter on the machine was not set after the last tool replacement.
Why 4: Why was the counter not set? The second-shift setup operator was not trained on the counter reset procedure. First-shift operators had been handling this task, and the procedure was passed informally rather than documented.
Why 5: Why was the procedure not documented? The tool change counter reset was not included in the setup work instruction because the instruction was written before the counter system was installed.
Root Cause
The setup work instruction did not include the tool life counter reset procedure, and no formal training existed for the task. The work instruction had not been updated after the counter system was installed.
Corrective Action
Revise the setup work instruction to include the counter reset step with visual aids. Retrain all operators on all shifts. Add counter verification to the startup checklist. Update the PFMEA to reflect the tool life management risk and the new control.
Case Study 2: Repeat Customer Complaint on Cosmetic Defects
The Problem
A plastic injection molding supplier receives its fourth customer complaint in three months for surface blemishes (sink marks) on a Class A visible housing. Previous corrective actions focused on adjusting hold pressure and cooling time. Each time, the problem improved temporarily and then returned.
Containment
100% visual inspection with boundary samples is added at the end of the molding cell. Suspect inventory at the customer is sorted. The supplier provides replacement parts from a known-good lot.
Investigation
Because previous 5 Whys investigations led to the same answers (process parameter drift), the team uses a fishbone diagram to broaden the investigation across all 6M categories.
The fishbone reveals a factor not previously considered under “Material”: the incoming resin supplier had changed sub-tier sources six months ago, and the new resin lot shows slightly different melt flow properties. A scatter diagram confirms a correlation between incoming resin melt flow index (MFI) and sink mark rejection rate. Lots with higher MFI produce more sink marks.
Root Cause
The incoming resin specification did not include a melt flow index requirement. When the resin supplier changed sub-tier sources, the material properties shifted enough to affect part quality, but no incoming inspection criterion existed to detect the change.
Corrective Action
Add a melt flow index specification to the incoming material requirements. Implement MFI testing on incoming resin lots. Notify the resin supplier of the tightened specification. Adjust process parameters to optimize for the current resin and document the validated settings. Update the PFMEA and control plan to reflect the material risk.
Case Study 3: Recurring Machine Breakdown
The Problem
A welding robot on a high-volume assembly line experiences four unplanned stoppages in one month due to wire feed jams. Each stoppage causes 30 to 45 minutes of downtime. Maintenance clears the jam each time, and the line restarts. The problem keeps coming back.
Containment
No customer quality containment is needed (the issue is caught internally), but production scheduling adds buffer time to account for potential stoppages while the investigation proceeds.
Investigation
The team applies 5 Whys with the maintenance technician and the robot programmer.
Why 1: Why is the wire jamming? The wire is bird-nesting (tangling) at the feed roller.
Why 2: Why is the wire tangling? The feed roller tension is inconsistent, sometimes too loose to drive the wire smoothly.
Why 3: Why is the tension inconsistent? The tension spring on the feed assembly has lost its specified preload.
Why 4: Why has the spring lost its preload? The spring has not been replaced since the robot was installed three years ago, and it has fatigued beyond its service life.
Why 5: Why was the spring not replaced? The wire feed assembly spring is not included in the robot’s preventive maintenance (PM) schedule. The PM covers the torch, cables, and teach pendant but not the feed mechanism internals.
Root Cause
The preventive maintenance schedule for the welding robot does not include replacement of the wire feed assembly spring, which has a finite service life.
Corrective Action
Replace the fatigued spring immediately. Add wire feed assembly spring replacement to the PM schedule at an interval consistent with the manufacturer’s recommendation. Review all similar welding robots to verify their PM schedules include this component. Add a feed tension check to the weekly operator maintenance checklist as a secondary control.
Case Study 4: Incoming Material Failure
The Problem
A stamping operation experiences a sudden increase in cracking during a forming operation on a steel bracket. The rejection rate jumps from a baseline of 0.3% to 4.8% over a single production day. No process parameters were changed. The die, press, and lubrication system are all confirmed within normal operating conditions.
Containment
Production is stopped on the affected part number. All parts produced from the current coil are quarantined. A 100% visual and dimensional sort separates conforming from nonconforming parts. The previous coil’s production is verified as conforming through a sample review.
Investigation
Since no process parameters changed, the investigation focuses on the material. The team reviews the Material Test Report (MTR) for the current coil and compares it to previous coils.
The MTR shows that the current coil has a yield strength of 380 MPa, compared to the typical range of 280 to 320 MPa for previous coils. The elongation value is 18%, compared to a typical 28 to 32%. The material is harder and less ductile than expected.
A 5 Whys follows: Why is the material out of the expected range? Because the incoming material specification only requires compliance with the base steel grade (which has a wide allowable range), not tighter mechanical property limits needed for this forming operation. The supplier shipped material that met the purchase order specification but was unsuitable for the application.
Root Cause
The incoming material specification did not include application-specific mechanical property limits (yield strength and elongation ranges) required for successful forming. The generic steel grade specification allowed material properties that are technically compliant but functionally unsuitable.
Corrective Action
Revise the material specification to include tighter yield strength and elongation requirements based on the forming process capability. Communicate the updated specification to the steel supplier. Add mechanical property verification to the incoming inspection plan. Review the PFMEA to add material property variation as a potential failure cause and include the new incoming control.
Case Study 5: Internal Audit Finding on Document Control
The Problem
During an internal audit of the machining department, the auditor finds that the work instruction at Machine 12 is revision B, while the controlled document system shows the current revision is D. The operator has been following an outdated procedure for an unknown period. Revision C added a deburring step. Revision D changed the in-process inspection frequency.
Containment
The outdated work instruction is immediately replaced with the current revision D. The operator is briefed on the changes in revisions C and D. Parts produced since the last confirmed use of the correct revision are reviewed. The deburring step (added in revision C) is verified on a sample of recent parts. Those missing the deburring step are reworked.
Investigation
The team investigates why the document at the workstation was two revisions behind.
Why 1: Why was the wrong revision at the workstation? Because the document was not updated when revisions C and D were issued.
Why 2: Why were the revisions not distributed to this workstation? Because the document control coordinator distributed revisions via email to supervisors, and this supervisor did not print and replace the document at the workstation.
Why 3: Why did the supervisor not replace the document? Because there is no confirmation step in the distribution process. The coordinator sends the document but does not verify it was received and posted.
Why 4: Why is there no confirmation step? Because the document control procedure relies on a push-based distribution model with no verification loop.
Root Cause
The document control distribution process does not include a verification step to confirm that updated documents have been received and posted at the workstation. The process relies entirely on email distribution to supervisors without a feedback loop.
Corrective Action
Implement a document distribution acknowledgment process. When a revision is issued, the coordinator sends the document and the supervisor must return a signed acknowledgment confirming the old revision has been removed and the new revision is posted. Add a quarterly spot-check of workstation documents to the internal audit schedule. Evaluate transitioning to a digital document display system at workstations to eliminate paper-based distribution risks. Audit all current workstation documents to confirm they match the controlled document system.
Key Takeaways Across All Five Cases
Surface answers are not root causes. “Operator error,” “tool wear,” and “bad material” are starting points, not conclusions. Every case required digging deeper to find the systemic gap.
The root cause is almost always a system gap. A missing procedure step, an incomplete PM schedule, an inadequate material specification, a distribution process without verification. These are the causes that, once fixed, prevent recurrence.
Detection gaps matter as much as occurrence causes. In every case, the existing controls failed to catch the problem before it escalated. Addressing why the problem was not detected is just as important as addressing why it occurred.
The right tool depends on the problem. Simple linear chains use 5 Whys effectively. Repeat problems with previous failed corrective actions need a broader tool like the fishbone. Data-driven validation with scatter diagrams and material test reports strengthens the investigation.
Frequently Asked Questions
A root cause analysis example walks through a real or realistic problem scenario showing the investigation method, the chain of cause and effect, the systemic root cause identified, and the corrective action implemented. The case studies above demonstrate this process across five common manufacturing situations.
The most common root causes are system-level gaps: missing or outdated procedures, incomplete preventive maintenance schedules, inadequate material specifications, and processes that rely on human memory without verification safeguards. Individual human error is almost never the true root cause.
You have found the root cause when correcting it would prevent the problem from recurring, when you can trace a logical cause-and-effect chain from the root cause to the observed problem, and when the cause is supported by evidence rather than assumption. A good test: read the chain backward using “therefore” and check if the logic holds.
Reopen the investigation and broaden the analysis. The original root cause determination was likely incomplete. Use a different tool (switch from 5 Whys to fishbone, for example) to explore cause categories that were not considered. Look for contributing factors and material or environmental variables that may have been overlooked.
Detailed enough to identify a specific systemic condition that can be corrected. The investigation should include evidence at each step (data, records, observations), not just opinions. The documentation should allow someone unfamiliar with the investigation to follow the logic from problem statement to root cause to corrective action.
Yes. The investigation structure (define the problem, contain the damage, analyze the cause, implement corrective action, verify effectiveness) applies to any industry. Replace the manufacturing-specific details with your process context and the methodology works the same way.
Stay Current on Automotive Quality
PPAP updates, GD&T guidance, and practical quality resources delivered to your inbox. No spam, just what quality professionals need to know.
Subscribe
Leave a Comment